Exclusive Interview with Kairui International Rotating CEO Zeng Cheng: It's No Longer "All AI Positions Are Hot" — AI Talent Competition Is Shifting from General Skills to Scenario Implementation

If you’re trading stocks, just look at research reports by Golden Qilin analysts—authoritative, professional, timely, and comprehensive. Help you uncover high-potential thematic opportunities!

（Source: Securities Times）

Recently, UBTECH (Unitree?) has once again pushed AI talent recruitment to the forefront by globally recruiting a “Chief Scientist of Embodied Intelligence” with an annual salary range of 15 million to 124 million yuan.

What is the current state of AI talent recruitment? What trends will emerge in the future? What pain points exist in the hiring ecosystem? In response, Zeng Cheng, the rotating CEO of Korn Ferry International, told Securities Times reporter in an exclusive interview, saying that high-salary recruitment above tens of millions of yuan is not the industry norm. It usually only appears in a small number of top-tier companies and often involves individual cases during specific time windows. This move precisely shows that competition for AI talent is shifting from general capabilities to scenario deployment. When embodied intelligence reaches a key turning point, what companies are competing for is not just the talent itself, but the few critical people who can truly drive technology into real-world deployment and define the future landscape.

She also expects that in 2026, the recruitment heat structure of the AI industry chain will continue to rise structurally. It will no longer be “all AI positions are hot,” but instead “the roles that should be hot will get even hotter, and those that shouldn’t will naturally cool down,” entering a new stage of “rational prosperity.”

Three types of AI jobs have clear salary premiums

Securities Times reporter: What recruitment trends have you observed in the current AI industry chain?

Zeng Cheng: Based on real-time monitoring from Korn Ferry International’s data platform, the current demand for the AI industry chain’s recruitment is indeed maintaining a strong momentum. And AI talent needs are showing three relatively clear changes: first, companies are increasing their investment significantly in core algorithm and model engineering talent, focusing on the optimization of vertical industry models and upgrades to multimodal capabilities. Roles such as large-model algorithm engineers, algorithm researchers, and engineers who can implement model deployment and performance optimization have long remained in high demand, and the hiring difficulty is also relatively high.

Second, as embodied intelligence and humanoid robots enter large-scale validation stages, related frontier roles quickly become hiring hotspots. For example, the VLA/L4/world model direction—embodied intelligence algorithm engineers, multimodal fusion algorithm experts, and talent in robot intelligent control. Previously these roles had relatively scattered demand, but now they have become a focus of companies’ competition for talent, and the salary premium is very evident.

Third, AI is penetrating more deeply into real-world industries, especially with the deployment of agents, which drives growth in demand for industry-application-side roles. Companies prefer more composite talent who understands both technology and business. For example, agent development engineers, AI solution architects, and so on. At the same time, AI product managers and product solution experts who can convert technology into commercial value and deliver precise insights into user needs across different scenarios are also becoming scarce critical roles in the market.

In addition, as AI applications in enterprise core business scenarios become deeper and deeper, companies’ emphasis on model reliability, data quality, and business security has also increased noticeably. This, in turn, continues to drive sustained rises in demand for roles such as data governance, AI security assessment, and compliance review.

Securities Times reporter: Has the compensation level for AI industry chain recruitment seen a clear increase?

Zeng Cheng: Overall, the compensation level across the AI industry chain is not rising across the board. The core increases are concentrated in scarce tracks and core roles. For outstanding talent who job-hop, salary increase ranges generally focus on 20%—30%. For enterprises, they show greater salary flexibility for key technologies and leading positions.

The roles with truly clear premiums mainly fall into three categories: first, multimodal and embodied intelligence—especially composite talent that combines algorithm, system, and control capabilities. The salary premium for these core roles is significant. For example, the annual salary for senior experts in large-model algorithms is in the range of 1 million to 2 million yuan; and the annual salary for AI Agent technical senior engineers is between 400,000 and 700,000 yuan.

Second, model engineering and large-scale deployment. Simply put, these are engineers who can take models from the lab and actually bring them into real business, operate stably in production. Demand for this talent is strong, and salary growth is also especially prominent.

Third, composite roles of “technology + industry + product,” such as AI product managers and solution architects. This talent needs to understand technology, understand industry and business as well, and also be able to connect to commercial needs. Their compensation levels are also continuously rising. For example, the salary for senior AI product managers can reach 800,000 to 1 million yuan.

AI recruitment heat remains high, but the growth rate is stabilizing

Securities Times reporter: Do you think the recruitment heat for the AI industry chain in 2026 will continue, level off, or cool down? What is the basis for your judgment?

Zeng Cheng: I believe the recruitment heat of the AI industry chain in 2026 will continue to persist structurally. Overall it will remain high, but the growth rate will stabilize, and it is expected to enter a new stage of “rational prosperity.” Whether in China or in major global economies, AI has already been placed in a position of core competitiveness. Continued investment of policy, capital, and industrial resources determines that this will not be a hotspot limited to a short cycle. From the perspective of technology itself, artificial intelligence is still in the early stage of generational evolution. Multimodal large models, embodied intelligence, and AI for Science have already achieved some preliminary results, but there is still a long road to truly mature. As long as the technology continues to evolve rapidly, demand for high-quality talent will not stop.

At the same time, AI is accelerating its penetration across all industries. In the past it mainly concentrated in highly digitized areas such as internet and finance, but now it is accelerating its infiltration into real industries such as manufacturing, energy, agriculture, and healthcare. Every traditional industry’s upgrade toward digitization and intelligence will create continuing and stable talent demand.

But from the trend perspective, in the future it won’t be that “all AI jobs are hot.” Instead, it will be “the roles that should be hot will be hotter, and the roles that shouldn’t will naturally cool down.” For enterprises and talent, this is actually a good thing.

Securities Times reporter: That’s good for both enterprises and talent. How should we understand it? How would you evaluate the current AI industry chain recruitment ecosystem?

Zeng Cheng: I believe the current AI industry chain recruitment ecosystem is shifting from the early phase of high heat and strong emotions into a more rational, and more structural, stage. On the one hand, talent demand is starting to return to a value orientation. In the past, the market really did have a situation where “if it has anything to do with AI, people are being hired.” But now companies are becoming clearer that what truly determines competitiveness is not the number of roles, but whether talent can support business deployment. This change is pushing recruitment from “competing on buzzwords” to “competing on capabilities,” which is a necessary course correction for the entire industry.

Talent structure is being upgraded, and composite capabilities are becoming the mainstream direction. Companies are far less likely now to hire for a single point—either only algorithms or only business knowledge. Instead, they need composite talent who not only understands technical principles, but can also connect to industry scenarios, and has product awareness. In a sense, this is also pushing talent to evolve from the traditional “T-shaped” structure into a multidimensional “U-shaped/兀-shaped” structure. This is a long-term positive for improving overall AI talent quality.

Agile employment forms are changing from being supplementary options to strategic tools. This is something we have observed very clearly in the past two years. With AI technology iterating faster, it’s difficult for companies to cover all high-end capability needs through traditional headcount systems. As a result, more and more companies are introducing key capabilities through project-based experts, independent consultants, and other methods. This model, on the one hand, reduces companies’ labor costs and trial-and-error risk; on the other hand, it also provides senior expert talent with more flexible and diverse career paths. Taking one cross-industry company we serve that has entered the AI industry chain as an example: based on our deep understanding of the founder’s entered track, including his own background, we used business and organizational diagnostics to help the founder clarify business development directions and key talent needs. Instead of saying we would attract and “headhunt” industry top talent according to conventional approaches—which would not fit the track and the company’s real circumstances in terms of time cycle and cost—we broke down key modules such as product design, R&D, supply chain, and overseas marketing into project tasks. This supports the founder to rapidly assemble a cross-domain expert team within 3 months, forming an agile organization of “core founder + external expert network,” greatly shortening the product development cycle. The product is now set to be among the first to launch in overseas markets, realizing a breakthrough from 0 to 1.

Recommendations: shift “hiring by grabbing” to balancing “developing talent + employing talent”

Securities Times reporter: In a more rational and more structured AI recruitment ecosystem, are there also risks that need attention?

Zeng Cheng: I think the current recruitment ecosystem is indeed becoming more rational, but there are also certain risks that should be kept in mind. First, high-end talent is overly concentrated, making it hard for small and medium-sized enterprises to “find the right people.” Top AI talent is monopolized by leading companies and star startups. This increases the difficulty for small and medium-sized enterprises to acquire talent, which may weaken the industry’s overall innovation vitality to some extent, and even form a “top players dominate” pattern.

Second, companies prefer “plug-and-play,” compressing growth space for junior talent. Many enterprises show clear preference during recruitment for senior talent with more than 8 years of experience, while investment in junior talent with 1 to 3 years of experience is insufficient. At the same time, some companies lack a well-developed talent cultivation system. After hiring talent, they are unable to provide an appropriate development platform, leading to consistently high talent attrition rates. If systemic cultivation mechanisms are lacking over the long term, a talent gap may emerge in the future.

Third, a short-term profit-driven mindset is rising, bringing risks of misallocation of resources. Some enterprises and individuals overly focus on short-term compensation returns while ignoring long-term capability building and business value creation. Once the market environment changes, it is easy to fall into a situation of “high costs, low output.”

Securities Times reporter: Given this situation, what would you recommend?

Zeng Cheng: For the industry ecosystem, I suggest establishing a more open talent mobility mechanism—encourage talent from big companies to move into small and medium-sized enterprises and traditional industries. Through models such as talent sharing and technical advisory, AI capabilities can be more widely empowered to serve the real economy. For companies, I suggest shifting from “hiring by grabbing” to balancing “developing talent + employing talent.” On the one hand, quickly acquire scarce capabilities through flexible employment arrangements, independent consultants, and similar methods. On the other hand, increase internal training investment, build a composite talent cultivation system for “AI + business.” At the same time, do reverse validation: define roles using real business problems, maintain rational hiring, and improve talent development and retention systems.

Companies must think through their needs before they start recruiting. Many companies’ biggest misconception is: “seeing that others are hiring, I also need to hire.” But they haven’t figured out what this position is actually meant to solve. Is technology blocking progress? Does the product need a breakthrough? Or are they already at a critical stage for business deployment? If you haven’t figured this out, even if you hire people, it will very easily turn into a situation where “people are expensive, but nobody knows what they’re supposed to do.”

High-end talent doesn’t necessarily need to be “bought out” right from the start. For extremely scarce talent at a very high level, it’s entirely possible to cooperate through a project-based or advisory-based approach for a period first. This not only verifies capability and fit, but also reduces the risk of the company making a one-time large investment. While companies are疯狂抢 mature talent, they also need to build a mechanism for identifying high-potential talent. Some people may not be able to “fight hard battles” immediately, but they may have strong learning ability, good systems thinking, and genuine enthusiasm for both technology and business. Once given the right environment, their growth speed often exceeds expectations.

For talent, you should build a “Π-shaped” capability structure. You must have a sufficiently deep technical vertical axis—such as one direction within algorithms, systems, or engineering. At the same time, you need to understand industry, business, and product horizontally, knowing what the technology is ultimately used to solve. Single-point capabilities are easier to replace, but connection capabilities will become more and more valuable. Meanwhile, keep a balance between hands-on execution and thinking—being able to get into the trenches to write code and run experiments, and also being able to step out of pure technical work to think about industry trends, user value, and the underlying commercial essence.

A massive amount of information and precise analysis—exclusively on the Sina Finance APP